dsdreamer wrote: > Clive, I did some experiments with "-u vM::1.5" and I consistently get a > tiny gap, or click sound at 10 seconds before each track ends using 40ms > ALSA buffer lengths. If I have just one track in the playlist then this > doesn't happen. It seems like I get the click at the point where > buffering up the start of the next track happens. It is very strange > that I don't hear this with "-u vL::1.5". I also don't hear it with 80ms > of ALSA buffer length "-a 80:::". > > To summarize, it is 100% repeatable that if I use "-u vM::1.5" with "-a > 40:::" buffer, I will hear a pop noise at between 10 and 9 seconds of > remaining track time, whereas "-u vL::1.5" sounds clean with the same > ALSA buffer length. This is with standard 44.1kHz 16 bit FLAC rips, and > running SBS on the same Wandboard, with an NFS mounted music library. >
OK, bottom line, the buffer needs to be increased from 40 if you are upsampling. I'm probably not going to do a very good job of explaining this, but basically what is happening, is that 10-15 seconds before the end of one track, the next starts being buffered. When the data for that next track is fed into the resample library, CPU spikes for a fraction of a second. If you are running with a "small" alsa buffer, that CPU spike is probably enough to cause a glitch. As you have found, just increase the alsa buffer and it isn't a problem. The strange thing is, not sure why you are seeing this with minimum phase filter but not linear. I'd expect you to see it with both. Perhaps minimum phase uses more CPU. I haven't compared..... ------------------------------------------------------------------------ JackOfAll's Profile: http://forums.slimdevices.com/member.php?userid=3069 View this thread: http://forums.slimdevices.com/showthread.php?t=98544 _______________________________________________ unix mailing list [email protected] http://lists.slimdevices.com/mailman/listinfo/unix
